INDEX

Explanations

brand names and product-related terms

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 29-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L29R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.29.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 âĸĳ

-0.18

cheid

-0.17

ÏĦÏģÎ¿

-0.15

ambique

-0.15

Nag

-0.15

 Cherokee

-0.15

 parach

-0.15

 Cleveland

-0.14

 Spartan

-0.14

lyn

-0.14

POSITIVE LOGITS

 Mate

0.42

Mate

0.34

 Huawei

0.34

Kir

0.30

 mate

0.30

 Nova

0.28

mate

0.28

 nova

0.28

 Honor

0.28

nova

0.26

Activations Density 0.013%

brand names and product-related terms

No Comments

No Known Activations